Ask Your Question

Is it possible to extract a 3D Cad model from pictures like 123D?

asked 2012-11-12 17:39:59 -0600

jonjacson gravatar image

Autodesk has software able to create a 3D Cad model using a bunch of pictures, the point is that you send them the pictures and their software gets the cad model, I would like to understand what kind of mathematical operations are "behind the scenes".

Is it possible to use Open Cv to extract a Cad model from pictures?

I assume it is difficult any guidelines about where to start books, tutorials... would be very interesting, anything useful to get 3d information from pictures is interesting for me.

I have a science background with a bachelor on physics but I don´t know too much about this specific topic, what I mean is that I think I would be able to understand a technical paper talking about linear operators, matrix, geometric transformations and so on.

Thank you for your attention!

edit retag flag offensive close merge delete

2 answers

Sort by » oldest newest most voted

answered 2012-11-13 02:56:30 -0600

Ben gravatar image

There is a fascinating paper by David Nister: An Efficient Solution to the Five Point Relative Pose problem

It is some time ago that I read it, but if I remember correctly, Nister could retreive 3d informations from a handheld camera in real time by comparing subsequent frames, finding matches between the frames and using them to compute the relative pose between the two frames. It's hard stuff to read, but it could be a start for you. More recent papers might have cited him, so maybe that's helping you, too.

edit flag offensive delete link more

answered 2012-11-13 05:07:49 -0600

Simone gravatar image


the answer by @Ben is a good direction to search. I suggest you also take a look to this kind of systems (and the relevant pubblications cited in the website

  • this talk will introduce you to the topic and to the various (actually, a lot) possible approaches.
  • PMVS2 this is a 3D reconstruction tool from multiple images, it provides a sparse/dense reconstruction on the scene.
  • VisualSFM it's basically very similar to the above but a lot improved with gpu.

They both give you as output points. I guess you prefer to have some sort of geometric surfaces fitted to that data, which may be the starting point for you. Also it depends a lot on the prior knowledge you have about the shapes you want to recover: if you know that the objects are man-made with simple geometric shapes like then you can set and impose some constraints during the 3D reconstruction so that you get as output geometric primitives (box planes and so on) describing your object. For example this may be a first point to start.

Hope it helps


edit flag offensive delete link more

Question Tools


Asked: 2012-11-12 17:39:59 -0600

Seen: 967 times

Last updated: Nov 13 '12